The AI Classroom: How Generative AI Is Reshaping Education

A national survey of educators, students, and administrators reveals widespread AI adoption, persistent integrity concerns, and a system struggling to find consensus on how—or whether—to use the technology.

Student Usage of AI

Multiple surveys indicate that generative AI use is now the norm for a majority of students across K-12 and higher education.

Higher Education

A July survey by Inside Higher Ed and the Generation Lab found that approximately 85% of undergraduates use AI for coursework, including brainstorming, outlining, and studying. About 19% reported using AI to write full essays, with over half expressing mixed feelings about its utility.

A separate survey of over 94,000 California State University (CSU) students, faculty, and staff found that more than half of students use AI regularly. In that poll:

• 64% of students reported a positive effect on their learning
• 35% reported a negative effect
• 80% stated they would not submit AI-generated work as their own

K-12

An NPR/Ipsos poll of 545 K-12 teachers found that a little over half reported students do not use AI in class at all, while about 2 in 5 teachers reported students use it at least once a week. Data cited by a Columbia University student indicated that a majority of U.S. teenagers currently use AI chatbots, with more than half utilizing them for academic work.

Student Perspectives

"I use AI to understand complex subjects, generate practice problems, and create code for data analyses—but I avoid using it for writing entire assignments."
— Anjali Tatini, pre-med sophomore, Duke University

"I use generative AI for proofreading and checking my work against course rubrics."
— Hannah Elder, pre-law junior, University of North Carolina

AI Detection Software: Deployment and Accuracy

The use of software designed to detect AI-generated text has become a major point of contention, with significant concerns raised about its reliability and impact on students.

Reliability and Accuracy

Research by Mike Perkins, an academic integrity researcher at British University Vietnam, found that popular AI detectors—including Turnitin, GPTZero, and Copyleaks—can misclassify both human-written and AI-generated text. Accuracy rates reportedly decrease when AI-generated text is modified to appear more human.

Turnitin's own guidance advises that scores of 20% or lower are less reliable and that its AI writing detection should not be the sole basis for adverse actions against students.

A Case Study in False Accusation

Ailsa Ostovitz, a 17-year-old junior at Eleanor Roosevelt High School in Maryland, was accused of using AI on three assignments after an AI detection program indicated a 30.76% probability of AI use on a writing assignment. Ostovitz denied using AI.

The school district, Prince George's County Public Schools (PGCPS), stated the teacher used the tool independently, as the district does not fund such software and advises educators against relying on it due to documented inaccuracies. Following a meeting with Ostovitz's mother, the teacher reportedly changed their belief about the accusation.

Ostovitz now pre-screens her assignments using multiple AI detection tools to avoid future accusations.

District Contracts and Usage

• Broward County Public Schools, Florida: Holds a three-year contract worth over $550,000 with Turnitin. The district states the tool facilitates conversation and feedback rather than grading, and acknowledges the tool's inaccuracies.

• Shaker Heights City School District, Ohio: Spends approximately $5,600 annually for GPTZero licenses for 27 teachers. Teacher John Grady uses the tool as an initial indicator; if a score exceeds 50%, he conducts further investigation.

Bias Concerns

Zi Shi, a Shaker Heights junior whose first language is Mandarin, reported that his writing style has been flagged by GPTZero.

Carrie Cofer, an English teacher in the Cleveland Metropolitan School District, reported that a chapter of her Ph.D. dissertation was flagged as 89-91% AI-written by GPTZero during an experiment.

Prevalence

A nationally representative poll by the Center for Democracy and Technology indicated that over 40% of 6th- to 12th-grade teachers utilized AI detection tools during the previous school year.

Institutional Policies and Diverse Approaches

Educational institutions and individual teachers have adopted varied strategies—ranging from complete prohibition to structured integration.

The Analog Approach

Chanea Bond, an English teacher at Southwest High School in Fort Worth, Texas, has implemented an almost entirely analog approach, requiring handwritten assignments and journaling. She stated this ensures students develop critical thinking and writing skills without AI assistance.

Structured Integration

Maximillian Milovidov, a freshman at Columbia University, described a course titled "Writing AI" that mandated the use of AI. Students initiated their own ideas and outlines, then submitted drafts to chatbots, recording suggestions and articulating their reasoning for accepting or rejecting them.

Varied Faculty Views

"Using generative AI for college essays hinders students' skill development."
— Dan Cryer, English professor, Johnson County Community College

"Generative AI is a powerful collaborative tool. I encourage students to use it responsibly for outlining and feedback."
— Leslie Clement, professor, Johnson C. Smith University

CSU System-Wide Contract

The California State University (CSU) system signed a partnership with OpenAI to provide ChatGPT Edu to its students, faculty, and staff. The initial $17 million no-bid contract was signed in 2024, followed by a renewal of $13 million per year for three additional years.

A petition co-authored by San Francisco State University professor Martha Kenney against renewing the contract cited environmental impact, use of copyrighted material, and educational concerns. CSU stated the petition does not reflect overall sentiment and that the AI advisory committee unanimously recommended renewal.

AI for Administrative and Special Education Tasks

AI is also being used by educators for tasks beyond direct instruction—with growing adoption in special education.

Special Education

A survey by the Center for Democracy and Technology (CDT) found that 57% of special education teachers used AI to develop Individualized Education Programs (IEPs) in the 2024-25 school year, up from 39% the previous year.

Teachers at Riverview Middle School in Bay Point, California, reported using AI tools to speed up IEP goal writing. Studies from the University of Virginia and the University of Central Florida indicate that AI can produce IEP goals of equal or higher quality compared to teachers working alone, when used appropriately.

The CDT raised concerns regarding data privacy, potential bias, and that about 15% of teachers surveyed rely entirely on AI for IEP development.

General Teacher Usage

The NPR/Ipsos poll found that 60% of K-12 teachers reported using AI for work tasks, primarily to save time. However, 63% of those said the time saved was two hours or less per week.

Impact on Trust and Core Skills

Surveys highlight deepening concerns about AI's effect on the educational environment.

Critical Thinking

In the NPR/Ipsos poll:

• 54% of K-12 teachers said AI makes it harder for students to learn critical thinking skills
• 55% said AI is mostly a shortcut for students to avoid work

Trust

Nearly 60% of teachers in the NPR/Ipsos poll said AI is eroding trust between students and teachers. As a result, about 40% have required more assignments to be done by hand or in class.

Lack of Guidance

About half of teachers in the NPR/Ipsos poll said their school has offered no guidance on AI, or they are unsure of the guidance. Only about 40% said their school offers AI-related professional development.

Public and Teacher Sentiment

Nearly 80% of K-12 teachers think schools should teach responsible use of AI.

In the CSU survey, approximately 65% of students and 59% of faculty expressed skepticism that AI is benefiting education overall. Large majorities of CSU respondents expressed concern about AI's impact on:

• Creativity: 83% of students, 82% of faculty
• Job security: 82% of students, 78% of faculty
• The environment: 80% of students, 84% of faculty